My laboratory pursues two different lines of work which deal with the molecular, biochemical, and physiological analysis of muscarinic acetylcholine and vasopressin receptors. These receptors are prototypical members of the superfamily of G protein-coupled receptors (GPCRs). (I) STRUCTURE-FUNCTION ANALYSIS OF GPCRs GPCRs form one of the largest protein families found in nature, and estimates are that about 50% of drugs in current clinical use act on specific GPCRs or on GPCR-dependent downstream signaling pathways. To understand how these receptors function at a molecular level, we have used different muscarinic acetylcholine and vasopressin receptors as model systems. To elucidate the structural changes involved in ligand-dependent GPCR activation, we recently developed a novel disulfide cross-linking strategy which offers the great advantage that intramolecular Cys-Cys cross-links are generated with the receptor present in its native membrane environment (in situ!) (Ward et al., JBC 277, 2247, 2002). This approach is based on the ability of two cysteine residues that are adjacent to each other in the three-dimensional structure of a protein to form a disulfide bond, either spontaneously or under oxidizing conditions. Disulfide cross-linking experiments revealed that agonist activation of the M3 muscarinic receptor is associated with striking structural changes on the intracellular surface of the receptor protein (Ward et al., JBC 277, 2247, 2002). Using an analogous approach, we recently provided evidence that the relative orientations of the transmembrane helices of the M3 receptor differs from that of rhodopsin, the prototypical class I GPCR (Hamdan et al., Biochemistry 41, 7647, 2002). In addition, we recently initiated a novel project that involves the expression of various muscarinic and vasopressin receptor subtypes as well as different G protein alpha subunits in yeast (S. cerevisiae). A great advantage of the yeast expression system is that powerful genetic approaches can be applied to study GPCR structure-function relationships, allowing the screening of large numbers of mutant GPCRs or G proteins in a very efficient manner. Yeast strains were genetically modified in a fashion such that yeast growth was strictly dependent on productive GPCR/G protein coupling. We first demonstrated that M3 muscarinic and V2 vasopressin receptors retain proper G protein coupling selectivity when expressed in yeast (Erlenbach et al., J. Neurochem. 77, 1327, 2001; Erlenbach et al., JBC 276, 29382, 2001). The two receptors were then subjected to random mutagenesis to generate large libraries of mutant receptor clones. These mutant receptor libraries are currently being screened in yeast in order to isolate mutant receptors with novel/altered functional properties. Using this strategy, we recently isolated a series of mutant V2 vasopressin receptors which displayed novel G protein coupling profiles (Erlenbach et al., JBC 276, 29382, 2001). Systematic application of this approach (receptor random mutagenesis followed by yeast genetic screens), which does not rely on preconceived models of GPCR function, should greatly enhance our knowledge about the structural determinants governing GPCR function. (II) GENERATION AND ANALYSIS OF MUSCARINIC ACETYLCHOLINE RECEPTOR KNOCKOUT MICE Most of the important physiological actions of acetylcholine (ACh) are mediated by the binding of ACh to a group of cell surface receptors referred to as muscarinic ACh receptors (mAChRs). The mAChR family consists of five molecularly distinct receptor subtypes (M1-M5) which are abundantly expressed in most tissues or organs. However, primarily due to the lack of receptor subtype-selective ligands, the precise physiological and pathophysiological roles of the individual mAChRs have remained obscure. To address this issue, we, in collaboration with Chuxia Deng's lab at NIDDK, used gene targeting technology to generate M1-M5 receptor-deficient mice (KO mice). The M1-M5 mAChR KO mice were then subjected to a battery of physiological, pharmacological, behavioral, biochemical, and neurochemical tests. This analysis showed that each of the analyzed mAChR KO lines displayed specific functional deficits, indicating that each mAChR subtype mediates distinct physiological functions. M1 receptor KO mice showed a pronounced increase in locomotor activity in a variety of behavioral tests (Miyakawa et al., J. Neurosci. 21, 5239, 2002). Moreover, M1 receptor KO mice did no longer display ACh-mediated hippocampal gamma oscillations, a rhythmic network activity predicted to be critically involved in central information processing including memory and learning (Fisahn et al., Neuron 33, 615, 2002). Analysis of M2 receptor KO mice revealed that the M2 muscarinic receptor mediates pronounced analgesic/antinociceptive effects in the CNS (at both spinal and supraspinal sites; Duttaroy et al., Mol. Pharmacol., in press) and at peripheral nociceptive nerve endings (Bernardini et al., J. Neurosci. 22:RC229, 1, 2002). In vitro neurotransmitter release studies indicated that presynaptic M2 receptors regulate the release of ACh and other neurotransmitters in the cerebral cortex and hippocampus and in various peripheral organs (Zhang et al., J. Neurosci. 22, 1709, 2002; Zhou et al., Naunyn-Schmiedeberg's Arch. Pharmacol. 365, 112, 2002). M3 receptor KO mice displayed a significant decrease in food intake, combined with reduced body weight, low serum insulin and leptin levels, and decreased total body fat mass (Yamada et al., Nature 410, 207, 2001). More detailed studies revealed the existence of a novel cholinergic pathway that promotes food intake via ACh-mediated activation of hypothalamic M3 receptors. Previous studies demonstrated that M4 receptor KO mice showed enhanced sensitivity to the stimulatory locomotor effects of selective D1 dopamine receptor agonists. This observation suggests that M4 receptors may function physiologically to mediate inhibition of D1 receptor-mediated locomotor stimulation, probably at the level of striatal projection neurons where both receptors are co-expressed at high levels. Studies with M2/M4 receptor double KO mice revealed that activation of the M4 receptor subtype (besides the predominant M2 receptor subtype) contributes to the profound analgesia observed after administration of centrally active muscarinic agonists (Duttaroy et al., Mol. Pharmacol., in press). In vitro neurotransmitter release studies revealed that the M4 subtype is the predominant muscarinic autoreceptor in the striatum and several peripheral organs (Zhang et al., J. Neurosci. 22, 1709, 2002; Trendelenburg et al., Br. J. Pharmacol., in press) and that activation of striatal M4 receptors plays a key role in facilitating dopamine release in the striatum (Zhang et al., J. Neurosci. 22, 6347, 2002). Surprisingly, ACh, a powerful dilator of most vascular beds, lost the ability to dilate cerebral arteries and arterioles in M5 receptor-deficient mice (Yamada et al., PNAS 98, 14096, 2001). The M5 mAChR therefore represents an attractive novel target for the treatment of various CNS (cerebrovascular) disorders including Alzheimer's disease and stroke. We also found that the rewarding effects of morphine and cocaine and the severity of drug-dependent withdrawal symptoms were substantially reduced in M5 receptor KO mice (Basile et al., PNAS 99, 11452, 2002). The M5 receptor can therefore also be considered an attractive novel target for drugs able to suppress drug-seeking behaviors. These new findings should be highly relevant for the development of novel muscarinic drugs useful for the treatment of several major pathophysiological conditions including Alzheimer's and Parkinson's disease.